Article Text
PDF
Statistics from Altmetric.com
There has been much discussion of the justifiability of strikes by healthcare workers, but comparatively little discussion of the political processes through which strikes occur. This article focuses on the Trade Union Act 2016, which currently governs strike ballots in the UK. This legislation has important implications for healthcare workers being balloted on strikes (or other forms of industrial action).
The article first explains the legal requirements for a strike mandate and illustrates how votes in strike ballots can be counterproductive, bringing about outcomes opposite of what the voter wanted. Second, it argues that the turnout threshold responsible for these surprising results is undemocratic, since it means that outcomes need not reflect the wishes of those balloted. Third, it suggests that this has consequences for how balloted workers ought to vote, if they want the process to be more democratic. In particular, I propose that those who are neutral or indifferent have reason to cast spoiled ballots, in order to negate the turnout threshold. Further, those who oppose a strike also have reasons—if they value democratic decision-making—to vote, even though this may be less effective than abstaining.
As I write, in August 2023, junior doctors in the UK are once again being balloted over strike action.1 While the global picture is more varied, strikes by medical professionals have historically been rare in the UK. When British doctors went on strike in 2012, over changes to pensions, it was the first time since 1975.2 Since then, however, industrial action has become more common. Some attribute this to an erosion of patient-centred professional values, which has made workers more prepared to strike in response to deteriorating pay and conditions.3 4 Whether or not this is the explanation, the increased prevalence of strikes has caused some concern.
It is often objected that industrial action by healthcare workers may harm to patients. In fact, empirical evidence is mixed and inconclusive.5 Much depends on the exact form that action takes, such as the length of the strike, extent of participation and implementation of contingency measures.5 6 Nonetheless the possible risk, and subsequent danger of public perception of risk, raises questions regarding the justifiability of industrial action. These issues have been much discussed elsewhere.6–14 It is not my aim, here, to adjudicate on these matters. This paper is not directly concerned with the ethics of participation in strikes, but rather with voting in strike ballots.
One might be excused for thinking that there is no issue here. It would be natural to assume that one should vote according to one’s beliefs about the justifiability of the proposed strike. That is, if one believes the strike justified, then one should vote in favour of striking—or, at least, one is permitted to do so. Conversely, if one thinks the strike unjustified, then it might be assumed that one should vote against it. And, perhaps, one is justified in abstaining if one is unsure or indifferent.
In fact, given current UK legislation concerning strike ballots, matters are not as simple as might be supposed. One’s vote can have surprising effects. Those who vote may need to consider the possible consequences, which may not be what they intend. While the details given here are specific to the UK, similar issues may arise elsewhere, depending on relevant legislation. The crucial factor is that, according to current UK legislation, a mandate for strike action depends not only on votes for and against the strike, but also on overall turnout in the ballot. This means that one’s vote may actually bring about an outcome opposite of what one desires. My first aim here is to increase awareness of this. Second, I aim to show that ethical questions arise, not only regarding participation in strikes, but also participation in strike ballots.
The Trade Union Act 2016
Strike ballots in the UK are governed by the Trade Union Act 2016.15 This legislation has been described as ‘highly authoritarian’ and ‘anti-liberal’.16 Indeed, it might also be considered undemocratic, if democracy includes a right to strike.1 This Act imposes a number of restrictions on strikes and other industrial action, including a requirement for at least 50% turnout in ballots for industrial action. This is supplemented by a further requirement for 40% support for action specific to important public services, including healthcare.2 Thus, it is more difficult for healthcare workers to achieve a legal mandate than it is for those in non-essential occupations. However, it is the turnout threshold that is a particular concern here. This requirement is arguably undemocratic, because it means that a mandate for strike action need not reflect the wishes of union members consulted in the ballot.17
Strike ballots must be conducted by post. Further, in order to produce a lawful mandate for action, they must satisfy two conditions. First, a majority of the union members balloted must vote. This includes votes for and against action, as well as ballots returned blank or spoiled. So long as their ballot is returned, this counts as a vote for purposes of the turnout threshold. Thus, when I refer to abstentions, I mean only unreturned ballots. Second, a majority of valid votes cast must support action. Both of these requirements can be seen as ways of ensuring that any strike mandate enjoys popular support from members. For instance, the turnout threshold prevents a small group of disgruntled workers from securing a mandate for action, simply because of widespread abstention. However, this requirement also produces some rather surprising consequences.
The turnout requirement means that abstention is sometimes more effective at blocking industrial action than voting against it. In fact, sometimes voting against action can result in a mandate that would not otherwise have been achieved. Suppose there is a union consisting of 101 members. If 50 of these members vote to strike, while the other 51 abstain, then the ballot fails to achieve a mandate (scenario 1 in table 1). Though a majority of votes cast were in favour of action, satisfying the second condition, the turnout threshold was not met, since fewer than half of the members participated in the ballot. However, matters would be different had one more member voted. This would then have satisfied the turnout requirement, creating a mandate for action. Further, this would be true regardless of how that individual voted. That is, had 1 of the 51 abstainers decided to vote against the strike, rather than abstaining, then their vote would nonetheless result in a mandate (illustrated by scenario 2 in table 1). Thus, voting against a strike may result in a mandate, allowing it to happen, whereas not voting (abstaining) may prevent action. This, I submit, is counterintuitive. It also violates the responsiveness that we normally expect from democratic decision procedures.
Summary of voting patterns5
The turnout requirement can also block action that is popular with members, as in the case where 50 out of 101 members support action, while none oppose it (scenario 1). On the other hand, non-essential workers could achieve a mandate for strikes with only 26 out of 101 voting in favour, provided 25 vote against and 50 abstain (scenario 3). This is all the more surprising since, intuitively, the democratic case for action is stronger in the first of these two scenarios (there are more people voting in favour of action and fewer voting against it). That the law grants a mandate in the latter case, but not the former, illustrates that the ballot procedure is undemocratic.
Admittedly, this last puzzle would not arise for workers designated ‘essential’, including those in healthcare. As noted above, strike ballots in these sectors also require 40% support. Even where this requirement is met, a ballot can still fail due to lack of turnout. For instance, if 41 members voted to strike, while the other 60 abstained (scenario 4), there would be no mandate. However, the 40% support requirement can block a strike ballot that has sufficient turnout overall, such as a case where 40 vote in favour, 11 against, and 50 abstain (scenario 5). But there would be a mandate if 41 vote to strike, 40 vote against, and 20 abstain (scenario 6). As one can see, scenario 6 differs from scenario 4 only in that more people vote against the strike. Once again, a mandate may result from votes against strike action.
Consequences
The various restrictions imposed by the 2016 Act mean that mandates for strike action do not simply depend on the votes of union members. There are cases, such as scenario 3, where a mandate is achieved despite relatively little popular support (only 26 of 101 members voting in favour). On the other hand, the turnout requirement sometimes blocks a mandate where there is fairly strong support for action, as in scenario 1 (where 50 members vote in favour and none against). Given this, it is sometimes unclear how members should vote, even assuming they know whether or not they support strike action. Some members may be left facing a dilemma.
Those who support a strike3:
For those who support a proposed strike, matters are straightforward. They should vote in favour of the strike. By doing so, they contribute towards meeting the turnout threshold and towards a potential majority in favour of strike action. Voting in this way cannot be counterproductive to their aims.
Those who are neutral:
There may be some workers who neither support nor oppose a strike. That is, they are neutral or indifferent. There may be various reasons for this. They may think the reasons in favour of striking exactly balance those against or, perhaps more likely, that the reasons on each side are incommensurable and on a par.18 Alternatively, it may simply be that they are uncertain about the merits of a proposed strike. This differs from the previous case, because uncertainty concerns the individual’s epistemic condition, rather than the balance of objective reasons.
Those who have no preference either way may wish to abstain. Indeed, some have argued that one has a duty to abstain from voting in certain circumstances, such as where one is ignorant or indifferent.19 20 However, abstention is ordinarily a neutral option. That is, it does not influence the vote either way; the outcome will be determined by those who do vote. But this is not the case when there is a turnout requirement. In this case, abstention can serve to block action that would be justified. Consider again a case like scenario 1. Suppose the 51 abstainers are all uncertain or indifferent regarding the merits of strike action. They may wish to recuse themselves from the decision, leaving it to other members who do have clear preferences. However, their abstentions may end up blocking a mandate, even though everyone with a preference supports the strike.
If those who are undecided want the decision to be taken by their colleagues, rather than being the result of anti-union legislation, then they can make this more likely by returning blank or spoiled ballots. Though these spoiled ballots do not influence the decision, either in favour of or against strikes, they still count towards meeting the turnout threshold. Thus, had one of the 51 abstainers in scenario 1 returned a spoiled ballot, then this would have resulted in a mandate (because there would be a turnout of 51, with 50 favouring the strike).
While I am sceptical of any general duty to vote, I am inclined to think that those who are neutral or indifferent have moral reasons (if not duties) to return spoiled ballots. As pointed out, this effectively removes the turnout threshold from the equation, meaning that the decision is made by union members. Though I cannot fully argue the case here, I believe that this is more democratic, since it means that the decision is responsive to how members vote. If this is right, then returning spoiled ballots may promote democratic decision-making. If democracy is valuable, then union members may have pro tanto reasons to do this.4
Those who oppose a strike:
Finally, we have those who oppose a strike. Again, they may have various reasons for this position. For instance, some may think that their cause is not justified, while others may think that the cause is justified but that strike action is unlikely to achieve anything or that the harm to patients would be disproportionate. For present purposes, these differences are of little significance, although I assume that those in question wish to prevent a strike, rather than merely declining to participate in one themselves. I am concerned with how they should vote.
As demonstrated above, voting against a strike can be counterproductive, where it results in meeting the turnout threshold but is not enough to change the majority favouring action. However, this is not always the case. In other circumstances, where turnout is likely to be high, but the contest between those who favour a strike and those who oppose it is expected to be close, then it would be sensible for opponents to vote against it, rather than to abstain. The problem is not that abstention is generally more effective than voting against a strike, but rather that these effects can be difficult to predict. Those who oppose the strike may end up regretting their actions, whether they abstain or vote against the strike.17
Opponents of strikes may be unsure what they should do, unless they know how others are likely to act. However, this uncertainty is about what is most effective in promoting their desired outcome, rather than about what they want. That is, it concerns strategy. It may be that, in some circumstances, abstention would be the most promising means to preventing a strike. However, what is effective is not necessarily morally justified. Though this is a case of strategic non-voting, criticisms similar to those levelled at strategic voting. Those who abstain, in the hope of preventing action by not reaching the turnout threshold, seek to exploit electoral rules to their advantage, in ways that are arguably unfair (since they favour one side over the other) and undemocratic (since outcomes are not the result of how people vote).
I argued that, if we value democratic decision-making, then those who are neutral or indifferent have reason to return spoiled ballots, so the outcome is determined by member votes, rather than the turnout requirement. If this is right, then the same reasoning also applies to those who oppose strikes. While abstaining might sometimes be a more effective strategy, exploiting the turnout threshold in this way is acting in an undemocratic manner. If we want the decision to be made democratically, by union members, then they have reason to vote in the ballot, even though this may be counterproductive to preventing a strike.
Again, this reason to vote may have to be balanced against other competing reasons, since democracy is not the only thing that we value. For instance, those who oppose strikes because they think the harm to patients is unjustified may be forced to choose between their commitment to democracy (which may involve what they regard as unjustified harm to patients) and their commitment to protecting patients (which may require undemocratic means to block strikes). This conflict between commitment to democratic procedures and a substantive outcome is an instance of what has come to be known as Wollheim’s paradox.21 It is not my aim to determine what they should do all things considered. However, I contend that, in so far as they value democracy, they have reasons to vote against a strike, even if abstaining might be more effective.
Conclusion
While there has been much discussion regarding the ethics of strikes, there has been comparatively little discussion of the rules governing strike ballots. Of course, this is partly down to context. The legislation that I have described is specific to the UK and was only introduced relatively recently. Nonetheless, this legislation has important consequences for healthcare workers in the UK, who have been balloted on industrial action several times recently and who may well be balloted again in the near future. It is important for them to understand the possible effects of their votes and to consider the ethics of voting, as well as the ethics of striking.
Ethics statements
Patient consent for publication
Ethics approval
Not applicable.
Acknowledgments
I thank members of Southampton UCU for various discussions of strikes and ballots; Colin Farrelly for assistance with literature; and an editor and reviewer from the journal for valuable suggestions. It should not be assumed that any of these individuals endorse the views in the paper.
References
Footnotes
Twitter @DrBenSaunders
Contributors I am sole author and guarantor of the article.
Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.
Competing interests None declared.
Provenance and peer review Not commissioned; externally peer reviewed.
↵Several authors have defended relevant claims. Corey Brettschneider argues that democracy is not merely procedural, but incorporates certain substantive rights and values.22 Gourevitch defends a right to strike.23
↵This provision, of course, raises questions as to which services are important or essential and how this is determined. Governments may seek to extend these terms, to limit strikes.24
↵The cases here each concern those who support or oppose a particular call for strike action, rather than supporting or opposing strikes in general. I thank an anonymous referee for prompting me to clarify this.
↵Admittedly, one might think that the reasons to promote democracy pale in comparison to other considerations, such as patient safety. As noted above though, risks to patients are unclear.5
↵It is assumed, in all cases, that there are 101 members and no spoiled ballots.